Stability and properties of hydrate phase assemblages with high Al content in real microstructure

Date This project started on 20 October 2016

Status This project is Ongoing


Ecole Polytechnique Fédérale de Lausanne, CH
Eidgenössische Materialprüfungs- und Forschungsanstalt, CH
Aarhus University, DK


Replacing cement clinker with supplementary cementitious materials (SCMs) such as fly ashes or calcined clays leads to changes in microstructure, hydrate assemblage and mechanical properties. Such blends have high Al2O3 and SiO2 contents but are low in CaO compared to Portland cements leading to a hydrate assemblage with less or no portlandite, to C-S-H with low Ca/Si and rich in aluminium and in some cases to the formation of strätlingite. The co-existence of portlandite and strätlingite has been observed experimentally although thermodynamically they are not expected to coexist as shown in Figure 1 below. As the reaction of cement clinker and SCMs proceeds at different speeds, hydrates such as portlandite become destabilised with time while others, e.g. strätlingite, become stable within a microstructure with limited availability of water and space.


Figure 1: Ternary phase diagram showing the effect of fly ash on the hydrate assemblage. The stability area of portlandite is indicated in yellow, the stability area of strätlingite in green.

In addition to portlandite, C-S-H and strätlingite, siliceous hydrogarnet can precipitate. Siliceous hydrogarnet is not well crystalline and difficult to detect by X-ray Diffraction or thermogravimetric analysis in hydrated PC and calcium sulfoaluminate (CSA) cements. The presence of iron seems to promote its formation. Minor amounts of siliceous hydrogarnet have been observed in hydrated PC, but it completely replaces AFm phases at higher temperature and in PC hydrated for several decades.

This project aims to investigate the hydrates, the pore solution and their changes with hydration in low CaO and high Al2O3 and/or SiO2 systems. The study will focus on

  • PC with metakaolin, where strätlingite formation will occur only after days – weeks
  • CSA cement containing belite, where initially ettringite and strätlingite are formed, while C-S-H precipitates later

The methodology of this study will be based on a multi-technique study of two contrasting systems: 

  1. On the one hand the effect of high alumina content obtained by addition of metakaolin (calcined clay) eventually with limestone.  In such systems the hydration of the clinker silicate phases occurs first so the matrix of the paste is still dominated by C-S-H phase as in Portland cements. At longer hydration time strätlingite will occur
  2. On the other hand systems where ye’elemite is the first phase to react as in BYF or CSA cements.  In these system ettringite and hydrates alumina are the first phases to form with strätlingite and possibly C-S-H occurring later.

The contrast between these two systems will make it possible to look at the effects of kinetics and microstructure in addition to thermodynamics.  To the extent possible systems with similar overall chemistry will be compared.

The other variables to be included in the study are temperature, iron content, alkalis and water to cement ratio. An array of well-established experimental techniques for microstructural characterisation will be used, these have been chosen to give a complete description of the phase assemblages and microstructure with data which can be compared with that available from thermodynamic modelling.  It will also be possible to gain information on the link with mechanical properties and durability.

The main objective will be to identify the kinetics of the different reactions and how these are affected by temperature. The information generated will help understand

  • the effect of reaction kinetics of “new cement types” on phase assemblage and microstructure in presence and absence of portlandite;
  • the role of kinetics on the hydrate assemblage on the hydrate assemblage;
  • the influence of the pore solution on kinetics and hydrates;

and create a basis for modelling the evolution of the microstructure of high Al, Si cements: hydrates, porosity, water content,…